Inundated Quality Systems and Personnel

This week, I will address the third of nine common CGMP Quality System shortcomings; Inundated Quality Systems and Personnel.

If your company is like most of the companies that I have encountered, the Quality Systems and Quality Personnel are pushed to their limits. Sometimes this is the result of working with a less than robust process and/or an unstable product, but often it is the result of an over extended Quality System. I once inspected a facility that had over 5,000 non-conformance investigations within one year. Can you image what it takes to adequately complete and manage that many investigations? It amazes me how a manufacturer can be content to tolerate such a large amount of investigations.

Yes, Quality Systems must have adequate product quality controls, but sometimes these controls can be so stringent and rigid that it becomes almost impossible to meet them consistently. An effective Quality Systems must be able to adequately control the process, product, and testing, while still provide sufficient flexibility to avoid redundant and/or needless investigations? Overly stringent requirements can quickly inundate a Quality System to where it becomes inefficient and non-functioning.

It is “the boy that cried wolf” syndrome. When everything becomes a crisis, then a real crisis may be lost in the busyness of the day. In biology, there is a term called “adaptation”, where organisms become desensitized to a regularly occurring event, stimulus, or condition. For example, if the fire alarm is constantly going off for no reason, then people start ignoring it and are not likely to react when the alarm goes off for a real fire.

One of those 5,000 investigations involved a detrimental product/container interaction that was not identified as such. I am convinced that this investigation was not given its due focus due to the overwhelming number of investigations being conducted on a daily basis.

To overcome this weakness, the Quality Unit needs to ensure procedures and processes have sufficient flexibility to allow for non-impacting variability and eliminate needless investigations. Monitoring parameters should be periodically adjusted to eliminate “background static” (i.e., non-essential alarms and/or notifications). Once the needless investigations and nuisance alarms have been eliminated, the Quality System can focus on addressing real issues and improving product / process robustness.

Un-Refined Quality Monitoring

This week, I will address the second of nine common CGMP Quality System shortcomings; Un-Refined Quality Monitoring.

When limited development information is available, Critical Quality Attribute (CQA) and specifications may have been established too loose or too tight. If set too tight, the CQA and/or specifications may be regularly exceeded, thus causing extensive investigations and/or product to be scraped, reworked, and/or released under questionable justifications. If the CQAs and specifications are set to loose, poor product quality may go undetected. Setting precise CQAs, specifications, and processing parameters can be difficult to achieve using limited small-scale and bench-scale data. During a typical technology transfer, a set of CQA parameters and specifications are inherited from the development group’s smaller-scale experiences. Often, these CQA parameters and specifications have not been exactly determined and tend to be a best guestimation. Critical processing parameters are established with the goal of meeting these CQAs and specifications. Phase III clinical lots are made using these processing parameters and specifications. If these clinical trials are successful, then the CQAs, processing parameters, and specifications become the gold standard for determining product quality. I have often heard people simply justify process deviations by stating specifications were met. This may be acceptable justification or it may not, depending upon the how the specifications were determined in the first place.

During commercial scale-up, process validation is typically completed in full-scale equipment using these processing parameters and specifications. Often, the limits of the processing ranges are not tested during process validation and sometimes not even tested with bench-scale studies.

CQA and specification ranges must be established to allow for inevitable processing and testing variability, but obtaining product quality and/or stability data at the range’s extremes is generally not obtained by the time of commercial manufacturing. Due to this limited processing/product knowledge at the start of commercial production and the learning curve associated with new processes, procedures, and assays, manufacturing management often experience difficulties with the product and/or process during the first year of production. As manufacturing works through these difficulties, significant process and product characterization knowledge is obtained. After several months of commercial manufacturing, process/testing variability typically settles into a more stable pattern.

To overcome this shortcoming, the Quality Unit should closely monitor all critical processing parameters using Statistical Process Control (SPC) to enhance their process characterization knowledge. To facilitate this endeavor, product reviews should be completed quarterly instead of annual for new production lines. Ideally, better product and process characterization should be completed prior to initiating commercial manufacturing with the goal of obtaining each critical processing parameter’s edge of failure.

Narrow-Focused Development Program

This week, I will address the first of nine common CGMP Quality System shortcomings; Narrow-Focused Development Program.

Current Pharmaceutical Quality Systems tend to inherit a process/product designed and developed by a completely different internal organization. The research and development groups are typically concentrating on the enormous task of finding a suitable molecule and developing it into a useful drug product.

Consequently, commercial manufacturing efficiency and robustness is not typically the primary objective on anybody's agenda. Due to the tremendous resource requirements to research and develop a product, extensive product/process characterization knowledge is not often obtained prior to commercial scale-up.

To compound the issue, commercial management is often not introduced to the new product/process until the initial stages of technology transfer. By that time, it may be difficult for value-added input to be provided.

With these development restrictions, the initial commercial scale-up of a product is often less than robust. The first several years of commercial manufacturing often involve working through the inherited product/process short-comings. This trial-by-fire experience is often clumsy and inefficient.

To overcome this shortcoming, the Quality Unit should take an early interest in the product development and characterization process and continue that investment through the life cycle of the product. The Quality Unit should view product reviews as an opportunity to provide continual product improvement and characterization input.

Typical Shortcomings of GMP Quality Systems

During my many years of reviewing and working with CGMP Quality Systems, I have encountered several reoccurring shortcomings. Although a given Quality System is not likely to have all of these shortcomings, most quality systems that I have evaluated have at least one of them.

Nine common CGMP Quality System shortcomings are:

1. Narrow-Focused Development Program
2. Unrefined Quality Monitoring
3. Inundated Quality Systems and Personnel
4. An “It’s Approved, Don’t Improve It” Mentality
5. Under Utilized Quality Tools
6. Poor Quality Leadership from Senior Management
7. Resistance to New Technological Advances
8. Meager Internal Auditing Program
9. Ineffective Training Programs

Over the course of the next nine week’s I will elaborate on each of these Quality System shortcomings and provide suggestions on how to overcome them.

Genzyme Warning Letter Update

I requested the Establishment Inspection Report (EIR) from the FDA, but I only received the Form FDA 483 with a letter that states,

“In order to help reduce processing time and costs, certain material has been deleted from the records furnished to you because preliminary review of the records indicated that the deleted information is not required to be publicly disclosed.”

FDA’s Field Management Directive 145 indicates EIRs are routinely released after an inspection is considered closed. Giving their response to my FOI request, I have to assume they don’t consider this inspection closed. This is further support by Genzyme’s 5/21/09 news release that states, “The FDA has begun the inspection process at the plant to determine if the deficiencies cited in its warning letter have been satisfactorily addressed. When the FDA issues its formal findings, Genzyme will provide an update.” (http://www.genzyme.com/corp/investors/GENZ%20PR-052109.asp)

The FDA 483 indicates that the inspection was lead by ORA and included CDER product specialist, which may account for the coordinated response (CR and Warning Letters delivered together).

My initial reaction to this coordinated response was surprise. I do not frequently see Warning Letters being issued as part of a Pre-Approval Inspection (PAI). In addition, the follow-up inspection to verify corrective actions prior to approval seems atypical.

It appears the FDA is stepping up their enforcement policies, which isn’t necessarily a bad thing. However, I am not sure about the severity of the action in this case, given there was no indication that adulterated product was found. It seems to me that the FDA should have completed an appropriate risk analysis before taking this position (see my blog posting, 5/20/09).

I still need to read the EIR before I can make any solid conclusions, but it looks like I will have to wait until the inspection is considered closed before I can get my hands on it.

In my previous GMP Notes, I neglected to include the link to Genzyme’s Feb 09 warning letter.

http://www.fda.gov/foi/warning_letters/s7123c.htm

Aseptic Processing: The Holy Grail?

The recent Genzyme warning letter depicts the focus and severity that regulatory bodies place on aseptic processing, but what are the real risks?

Is aseptic processing the greatest risk associate with sterile products, as is often implied by the governments?

The answer: it depends.

The risk associated with aseptic processing can only be properly assessed by looking at the big picture and should not be calculated strictly from a microbial perspective.

Although certain products are required to be sterile, we do not always manufacture them through sterile processing. Sometimes we use aseptic processing due to the sensitivity of the product and the near impossibility of keeping all microbes out of the processing stream. The definition for sterile is: “free from living germs or microorganisms”; while the definition for aseptic is: “free from the living germs of disease, fermentation, or putrefaction.” The subtle distinction between the two is: sterile means no bugs; aseptic means no potentially harmful bugs.

The definition of aseptic processing concedes that we can have bugs in our product. With this said, it is fair to conclude that everyday sterile units are leaving a manufacturing site somewhere in the world in a less than sterile condition. Oh my, isn’t this dangerous and aren’t we going to kill some? Again, it depends.

Even though certain body systems, areas, and/or organs are designed to be sterile (i.e., brain, blood, lungs), microbial infiltration into these areas does occur. As a defense to these intrusions, the body’s immune system leaps into action.

Consider the heroin junky in a back alley with a used syringe and tarnished spoon. He heats up the spoon to melt the heroin, uptakes the liquid heroin into the syringe, and promptly injects it into his veins. I think we can all agree that this process is not aseptic let alone sterile.

So, why aren’t all back-alley heroin addicts dying of septic shock? The answer is likely that the injection is small volume and their immune system is functioning as it was designed. It is reasonable to assume that junkies are routinely injecting themselves with contaminated product, but the injected microbes are not serious pathogens and/or their immune system is neutralizing the invaders before they become a serious issue. If these junkies were injecting large volumes into their blood stream, it is possible for their immune system to become overwhelmed and result in a hazardous outcome.

With this said, we can conclude that aseptic processing is not designed to completely eliminate risks associated with administering sterile products and some patients are receiving sterile products that are less than sterile.

So, what are the real risks? Is the real risk that we give someone an adulterated product (less than sterile) or that we give someone a product that overwhelms their natural defenses and causes them harm?

The real risks associated with aseptic processing are dependent on the product, route of administration, dose volume, dose frequency, patient’s immune system, and specific contaminating microorganism. Products that foster microbial growth and/or can not be fully subjected to a sterilization process carry greater risks (i.e., blood products); Intravenous injections carry greater risks than Intramuscular; Large volume parenterals carry greater risks than small volume parenterals; Multiple dose products carry greater risks than single dose products; Patients with compromised immune systems are at greater risks; and Product contamination with a pathogen carry greater risks than normal flora contaminations.

Obviously, our risks are lower when we have fewer microbes slipping through our processing cracks and therefore, putting less reliance on a body’s immune system. We should strive to control our aseptic processes as tightly as feasible, but we don’t need to over react when we find a crack in the process or a point-of-control deviation. When we encounter these situations, we need to do what 21st Century GMP and ICH guidelines tell us to do, complete a complete risk assessment by evaluating the type of contamination, extent of the contamination, and the product’s type, administration, volume, frequency, and target patient.

The principles of risk assessment require the specifics of a given situation to be assessed and the potential to cause harm to be assigned. When we treat aseptic processing as the “Holy Grail” of the medical manufacturing industry, we completely ignore these principles.

In conclusion, aseptic processing risks should be evaluated on a case-by-case basis and should encompass the complete picture and not just the microbiological perspective. We must recognize that microbial contamination are likely occurring everyday without detection and/or harm. Equally, we must acknowledge microbial contamination in sterile products have caused great harm and even killed people. We simply should not treat all situations the same by either over reacting or under reacting. With sterile products, the stakes can be high, but we shouldn’t automatically assume the worst-case outcome without completing a comprehensive investigation and an appropriate risk assessment.

Stop the Madness

Is your Quality System inundated and overwhelmed with investigations, documentation, and bureaucracy?

Does it seem that you spend your whole day jumping from one fire to another?

The best way to render a Quality System ineffective is to treat every issue equally.

When all things are treated equal, the Quality System can become ineffective at addressing critical issues because it is struggling to stay afloat in an ocean of documentation and investigations.

When everything is elevated to crisis level, people stop paying attention to anything. If the fire alarm in your office goes off everyday when there is no fire, then you start ignoring the alarm. The same applies to Quality Systems. It is very important for an effective Quality System to readily identify and separate critical issues from mere inconveniences. Inconveniences need to be addressed, but they don’t need to be addressed in the same manner and method as a potential product impacting issue.

Consider the triage process for medical units. They must quickly decide on which patients are in critical condition and which patients can wait for their care. When done correctly, all patients receive the appropriate attention with the appropriate urgency. Non-critical patients are not neglected nor are they treated before critical patients.

Quality Systems need to be monitored to evaluate their effectiveness at identifying and addressing critical issues promptly without neglecting inconveniences all together.

Of course, the best case scenario is to prevent non-conformance events from happening in the first place. We need to stop simply reacting to issues and inconveniences and start looking for better ways to prevent their future occurrence.

Over time, the Quality System and manufacturing process should be fine tuned to eliminate the background noise (inconveniences) so critical issues can be clearly identified and investigated in a timely manner.